Outer Limits 4x4 Board

Ive got a 93 vitara with a g16b 16v efi. Is is allegedly a SE416W type II (accoring to redbook VIN ref) but it has a single wire O2 sensor. this seems to contradict information here;

http://www.outerlimits4x4.com/ftopic138 ... sensor+16v

and here;

http://www.outerlimits4x4.com/ftopic943 ... sensor+16v

I'm tring to troubleshoot poor fuel economy and seem to have a sensor that is reading rich (>0.6v ave at 2500 rpm). before i replace the sensor i want to check which i should have.

I have the setup described by Joe in the 1st post and have confirmed that i have +12V and earth going to the male (ecu) side of the 4-pin plug for a heated 3 or 4 wire sensor.

Extractors were put on by the previous owner and the single black sensor wire appears to have been cut and re-joined at 3 points between the sensor and the plug.

Soooo.... Which sensor should i be running and will swapping to the heated sensor get my ecu running closed mode sooner and improve my economy? Is there a pinout on the ecu that will indicate closed or open mode operation?

so urs is the later G16b?

differences are easy to tell in the airbox wiring plug.

let me know what plug u got here and i should be able to tell u what its meant to have.

also these have early and late in there, u may be able to check urself easy.

http://www.suzukiinfo.com/vitara/vitara ... 1_mpfi.pdf
http://www.suzukiinfo.com/vitara/wiring ... g16a-b.pdf

i've also got a air/fuel mixture jaycar meter thing that shows the lean/rich.

if u want to borrow it i can arrange it easy enough i think. just undo a couple wires.

AFM plug is 5 wire
Left to right

red black X1
grey yellow X2
grey red X1
grey black X1

Going to borrow a scope to get an accurate reading on the sensor i have at the moment (I don' think my multimeter has the response time to correctly measure the V fluctuation when in closed loop)

Cheers

so going by your wire codes u have the early one same as mine.

i run a single wire one same as what came with my motor. i have a new spare single one if u want to test it also.

going by the diagrams also it says dashed line is for optional heated ones. i think these may have been sold in NZ and colder climates only.

if u find out anything with the scope, i'd luve to test mine also as suffereing from same issue it seems but could be down to a few issues, be good to sort one out.

going by my gauge it is correct as comes rich on idle and then drops off warming up then flicks up when booting it.

It might pay to check that you've got a good earth with a single wire sensor. On my old Feroza somehow the exhaust manifold wasn't earthed to the engine ; then, after I fixed that problem I discovered it was surging a bit which turned out to be the sensor dying. They respond too slowly when dying and the mixtures are all over the place.

You should be able to see the voltage variation OK on a multimeter. When you back off the acclerator (eg. going downhill) it should go lean and if you floor it it should go rich. Otherwise it will be oscillating. If you can see any delay in the voltage change then get a new sensor.

whats a good way to earth manifold? run sep earth to lug somewhere? or just test with continuity on multi is enough?

from what i can see on my led tester it goes rich under load/revs

going by my dodge-o-meter im leaning out at idle and rich at revs/load....
I'm already using an ex manifold stud as my earthing point for my second battery so good earth is not a drama.

i'm not 100% sure with suzuki, so someone can correct me if they know better but on most cars that i have had to play o2 sensors on, they run in open loop at idle and also on full throttle. so any checks at idle or full throttle will be irrelevant, hold your revs at around 2000 for a while (20 or 30 seconds) this should put it in closed loop and while maintaining your revs check your voltage. should be oscillating between 700-1100mv but i'm starting to stretch my memory with the figures
jason

There's all sort of things that need to be right to enter closed loop. Older vehicles (I'm stretching my memory too) won't go into it at idle. You also need to have the coolant up to temperature and have the throttle position within certain parameters. (ie. if you floor it don't expect closed loop, it should go rich.)

Here's something that might be worth reading http://carl.outerlimits4x4.com/ftopic24 ... gen+sensor I wrote it when I was having issues with my Feroza. It will still be relevant for your situation.

mr green wrote:i'm not 100% sure with suzuki, so someone can correct me if they know better but on most cars that i have had to play o2 sensors on, they run in open loop at idle and also on full throttle. so any checks at idle or full throttle will be irrelevant, hold your revs at around 2000 for a while (20 or 30 seconds) this should put it in closed loop and while maintaining your revs check your voltage. should be oscillating between 700-1100mv but i'm starting to stretch my memory with the figures
jason

yep my main measurements are taken at around 2500 rpm after a 5 min drive around my burb. if the engine enters open loop at idle ill start ignoring my measurements there. From what im finding about the bosch sensor im running when in closed loop it should cycle between .1 and .9 V at up to 8 times per s. The mean V (ie at stoich) should be 0.45 V. At 2500rpm i seem to spend more time above 0.45V (ie rich) but as i said my multimeter does not have the reolution give my definate results.

This seems to be in line with murcods data (cheers who'd have thought there would be useful data outside of hte zook section ). i just lack the tools to make the measurement at the required resolution

I should add that the 3 joins in the black wire that runs to the plug are all soldered ( i wonder if that has contributed to my probs as described by mightymouse)

So i borrowed a mates CRO and hooked it up. Confirmed I'm going into closed loop at revs (seems to transition from open to closed somewhere between 1000 and 1500 rpm). My Vmax was 768mV and my Vmin was 128mV. I could not get the cro to give me a mean or Vrms (stupid 100meg probe that wont slow down and measure a 1hz signal )
Pezooki, I figure you might have some info... if you read this . Any hints on tweeking a CRO to give a mean on such a slow cycling signal. Can i slow down the measurment resolution or something.

I did confirm my earlier observation that its spends more time in the part of the cycle near 768mV and only briefly dips down into the lean low voltages near 128mV.

I'm going to get the zook onto a full 5-gas analyser next and see what it says then replace the O2 sensor and repeat both the CRO and gas analyser.